Epstein Barr virus-directed T-cell therapy for refractory EBV-PTLD in a toddler post Orthotopic heart transplantation.

Epstein-Barr Virus (EBV) is a ubiquitous herpes type virus that is associated with post-transplant lymphoproliferative disorder (PTLD). Usual management includes reduction or cessation of immunosuppression and in some cases chemotherapy including rituximab. However, limited therapies are available if PTLD is refractory to rituximab. Several clinical trials have investigated the use of EBV-directed T cells in rituximab-refractory patients; however, data regarding response is scarce and inconclusive. Herein, we describe a patient with EBV-PTLD refractory to rituximab after orthotopic heart transplantation (OHT) requiring EBV-directed T-cell therapy. This article aims to highlight the unique and aggressive clinical presentation and progression of PTLD with utilization of EBV-directed T-cell therapy for management and associated pitfalls.

Propranolol Therapy in Tetralogy of Fallot: Treating the Echocardiogram or Treating the Patient.

Hypercyanotic spells are one of the defining clinical features of Tetralogy of Fallot (TOF). Limited data exist on peak Doppler right ventricular outflow tract (RVOT) gradient as a risk factor for the development of hypercyanotic spells, frequency of prophylactic use of propranolol based on peak RVOT gradient, and its impact on preventing the occurrence of hypercyanotic spells. We aimed to quantify peak RVOT gradients as measured on transthoracic echocardiography in infants with unrepaired TOF and assess for correlation with clinical symptoms of hypercyanotic spells. We also assessed the frequency of pre-operative use of propranolol, indication for medication initiation, and occurrence of hypercyanotic spells with or without propranolol use. Retrospective analysis was performed on patients at our institution who were born between February 1, 2011 and May 31, 2023. Patients were excluded if they were maintained on prostaglandin infusion or underwent palliative shunt placement or balloon valvuloplasty prior to complete surgical repair. Demographics, occurrence of hypercyanotic spells, propranolol use, peripheral oxygen saturation, age at surgical repair, and peak RVOT gradient at the time of propranolol initiation were collected from the electronic medical record. If no propranolol use was recorded, the single highest maximum RVOT gradient prior to surgery was collected. 203 patients were identified, of which 92 patients were included in analysis. Thirty-six (39%) patients received propranolol and 19% of patients developed hypercyanotic spells prior to surgery. Patients with higher peak RVOT gradients were more likely to be started on propranolol even in the absence of overt symptoms, and they also demonstrated more systemic desaturation. Additionally, peak RVOT gradient was found to be a poor predictor for the development of hypercyanotic spells. Wide clinical variation exists in the prophylactic use of propranolol for prevention of hypercyanotic spells. Peak RVOT gradient is not a reliable tool for prophylactic propranolol initiation to prevent hypercyanotic spells.

A multi-centre randomised control trial of directly observed therapy to promote medication adherence in paediatric heart transplant recipients.

Medication non-adherence remains a significant challenge for adolescent heart transplant recipients. Building on the success of a pilot intervention study, herein we describe the protocol for a follow-up randomised control trial using mobile video directly observed therapy, featuring several innovations, to promote medication adherence in a multi-centre sample of adolescent heart transplant patients.

Comprehensive Approach to the Management of Patients with HLHS: Analysis of 100 Consecutive Neonates.

BACKGROUND: We report our comprehensive approach to patients with hypoplastic left heart syndrome (HLHS) and describe our outcomes in 100 consecutive neonates. METHODS: One-hundred consecutive neonates (2015-2023) were stratified into 3 pathways: Pathway(1): 77/100=77% were standard-risk and underwent initial Norwood (Stage 1). Pathway(2): 10/100=10% were high-risk with noncardiac risk factors and underwent initial Hybrid Stage 1. Pathway(3): 13/100=13% were high-risk with cardiac risk factors: 10 underwent initial Hybrid Stage 1 + ventricular assist device insertion (HYBRID+VAD), while 3 underwent primary transplantation. RESULTS: One-year mortality=9/100=9%. Pathway(1): Operative Mortality for initial Norwood (Stage 1)=2/77=2.6%. Of 75 survivors of Norwood (Stage 1): 72 underwent successful Glenn, 2 underwent successful biventricular repair, and 1 underwent successful cardiac transplantation. Pathway(2): Operative Mortality for initial Hybrid Stage 1 without VAD=1/10=10%. Of 9 survivors of Hybrid (Stage 1): 4 underwent successful cardiac transplantation, 2 died while awaiting cardiac transplantation, 3 underwent Comprehensive Stage 2 (with 1 death), and 1 underwent successful biventricular repair. Pathway(3): Of 10 HYBRID+VAD: 7/10=70% underwent successful cardiac transplantation and are alive today and 3/10=30% died on VAD while awaiting transplantation. Median VAD support time=134 days (range=56-226). (Two of three patients who were bridged-to-transplant with prostaglandin underwent successful transplantation and one died while awaiting transplantation.) CONCLUSIONS: A comprehensive approach to the management of patients with HLHS is associated with Operative Mortality after Norwood of 2/77=2.6% and an overall one-year mortality of 9/100=9%. 10/100 patients=10% were stabilized with HYBRID+VAD while awaiting transplantation. VAD facilitates survival on the waiting list during prolonged wait times.

Benchmarking Transformer-Based Models for Identifying Social Determinants of Health in Clinical Notes.

Electronic health records (EHR) have been widely used in building machine learning models for health outcomes prediction. However, many EHR-based models are inherently biased due to lack of risk factors on social determinants of health (SDoH), which are responsible for up to 40% preventive deaths. As SDoH information is often captured in clinical notes, recent efforts have been made to extract such information from notes with natural language processing and append it to other structured data. In this work, we benchmark 7 pre-trained transformer-based models, including BERT, ALBERT, BioBERT, BioClinicalBERT, RoBERTa, ELECTRA, and RoBERTa-MIMIC-Trial, for recognizing SDoH terms using a previously annotated corpus of MIMIC-III clinical notes. Our study shows that BioClinicalBERT model performs best on F-1 scores (0.911, 0.923) under both strict and relaxed criteria. This work shows the promise of using transformer-based models for recognizing SDoH information from clinical notes.